All?gre Claude J. Isotope Geology

Подождите немного. Документ загружается.

is still referred to as juvenile water. Har mon C rai g (he again) studied geothermal water to

determine the isotope composition of any juven ile water. He showed that the d values of

geothe rmalwaters from the same source canbe plotte d on a (d D, d

O)diagram on straight

linespretty wellparallelto the d

O axis and which cutthe straightline of precipitation cor-

responding to the composition of rainwater for the region. And so the composition of

geothe rmalwater canbe explain ed by th e evolution ofmeteoricwater viaisotope exchange

ofoxygen with the country rock.There is no need to invoke juvenile water from the mantle

to explaintheseisotope c ompositions (Figure 7. 21).

As these relations are systematic for allthe geothermal regions studied, Craig concluded

thatthe inputof juvenile water intoth e currentwater cycle is negligible and thatgeothermal

waters are only recycled surface water. The same go es for water from volcanoes. This

hypothesis has been con¢rmed by more elaborate studies ofvariations in the isotope com-

position of geothermal water over time. In many cases, it has been shown that variations

trackedthose observed in thesame place for rainwater, withatime lagc orresponding to th e

transittimewhichvaried from monthstoyears.

EXAMPLE

Iceland’s geysers

In some instances, such as the geysers of Iceland, the straight line of (dD, d

O) correlation is

not horizontal but has a positive slope (Figure 7.22).

–50

–100

–150

–15 –10 –5 0 5

Wairakei

Laederello

The Geysers

Iceland

Niland

Hekla

Lassen Park

Steamboat Springs

δD

Figure 7.21 Correlation diagram for (

O, D/H) in geothermal waters. They form horizontal lines

cutting the meteoric straight line at the point corresponding to local rainwater. This is interpreted by

saying that water has exchanged its oxygen isotopically with the rock but the hydrogen of water does

not change because it is an inﬁnite hydrogen isotope reservoir compared with rocks that are relatively

poor in hydrogen. After Craig (1963).

A spa water company signed a research contract with a Parisian professor to study the isotopic

composition of the water it sold to prove it was ‘‘juvenile’’ water, a name whose advertising value can

be well imagined. As the studies showed the water was not juvenile, the company terminated the

contract and demanded that the results should not be published!

399 The isotope cycle of water

Suppose we begin with rainwater of local composition and that this water undergoes

distillation by evaporating. Then:

 d

O;D

þ 10

ð

 1Þln

O  d

O; O þ 10

ð

O  1Þln

Eliminating ln

gives:

 d

O;D

O  d

O; O





 1



 1



We know that at 100 8C, for water–vapor fractionation, 

¼1.028 and d

O ¼1.005. The

slope corresponds to 5.6, a lower value than that of equilibrium fractionation (8). The effect is

therefore a combination between exchange and distillation.

In fact, in nature, isotopic compositions of geothermal water or vapor are combinations

between Rayleigh distillation and the water–rock oxygen isotope exchange, between kinetic

fractionation and equilibrium fractionation. A horizontal slope indicates that isotope

exchange has been possible and so the transit time is long. When the slope is identical to

that of the Rayleigh law, the transit time has been short.

7.6 Oxygen isotopes in igneous processes

Examination shows thatthe

O isotope composition ofunaltered rockof deep origin,

whether o cean basalts or ultrabasic rocks, is extraordinarily constant at d

O ¼þ5.5

(Figure 7.23).This value is analogous to the meanvalue of meteorites. It has th erefore been

agreed that this value is the reference value for the mantle.When taking stockof measure-

ments on basic or acid, volcanic or plutonic igneous rocks, the results are found to divide

between:

-10

-20

-30 -20 -10 0 +10

δD (%)

O (‰)

Meteoric line

Geysers

Figure 7.22 Correlation diagram for (

O, D/H) in acidic geothermal waters and geysers. The diagram

is identical to the previous one except that these are acidic geothermal waters with a high sulfate content

whose pH is close to 3 and for which the correlated enrichment in D and

O results mostly from more

rapid evaporation of light molecules with kinetic fractionation into the bargain. After Craig (1963).

400 Stable isotope geochemistry



igneousrockswitha d

O valuegreater than5.5;



igneousrockswitha d

O valueless than 5.5, andsomewith negative values.

These twotrends correspondtotwo types ofphenomena a¡ecting igneous rocks: contami-

nationbycrustalrocks and postsolidus exchanges with hydrothermal £uids.

7.6.1 Contamination phenomena

These phenomena are classi¢ed under two types: those involving mixing at the magma

source where melting a¡ected both acidic and basic metamorphic rocks, and those

where contamination occur red when the mag ma was emplaced. The latter process,

known as assimilation, obeys a mechanism already accounted for by Bowen (1928).

Mineral crystallization in a magma chamber releases latent h eat of crystallization.

This latent heat melts rock around the edges of the magma chamber leading to their

assimil ation.

–5

Syenites and trachytes

Anorthosites

Ultramafic and rocks

Basalts and gabbros

Andesites

Granites,

granodiorites and tonalites

Ash flow and

rhyolitic tuffs

Rhyolites and dacites

Italian volcanic rocks

Border zone of

granitic bodies

Scottish Hebrides

Skaergaard intrusion

San Juan Mountains

Western Cascade Range

Western Nevada Au-Ag

Boulder batholith

Iceland

Eclogites

Lunar rocks

Meteorites

0+5+10 +15

–5 0 +5 +10 +15

O (‰)

Figure 7.23 Values of d

O in rocks and minerals. After Taylor (1974).

401 Oxygen isotopes in igneous processes

m # L ¼ m " C

DT;

where L is thelatentheat, m#the mass ofcrystalsprecipitatingper unittime, m"the massof

rock assimilated, C

thespeci¢cheatofthesurrounding rocks, and T thetemperaturedif-

ference between th e wall rock and the magma. If we can write m#¼kM,thenkML ¼m#

T, therefore:



The magma is contaminated isotopically toobythe mixing law:

ðd

OÞ

¼ðd

OÞ

magma

ð1  xÞþðd

OÞ

country rock

ðxÞ

with x ¼ m #=MðÞ, because the oxygen contents of the country rock and the magma are

almost identical. This was shown by Hugh Taylor (1968) of the California Institute of

Technology (see also Taylor, 1979).

Exercise

What is the d

O value of a basaltic magma whose d

O ¼0 and which assimilates 1%, 5%,

and 10% of the country rock whose d

O ¼þ20?

Answer

The contamination effect therefore increases the d

O value because sedimentary and

metamorphic rocks have positive d

O values. An interesting approach to studying the

contamination of magmas by continental crust is to cross the studies of oxygen isotopes

with those of strontium isotopes. The (O–Sr) isotope diagram can be calculated quite simply

because it is assumed that the oxygen content is analogous in the different rocks. The mixing

diagram depends only on the Sr contents of the two components of the mixture. Figure 7.24 is

the theoretical mixing diagram.

Such combined studies have been made of volcanic rocks of the Japan arcs and the

Peninsular Range batholith in California (Figure 7.25).

Exercise

A basaltic magma is emplaced and assimilates 1%, 5%, and 10% of the country rock. The d

values are those of the previous exercise. The

Sr/

Sr values are 0.703 for the magma and

0.730 for the country rock. The Sr content of the magma is 350 ppm and that of the country

rock is 100 ppm. Calculate the isotopic compositions of the mixture and plot the (d

Sr/

Sr) diagram.

Assimilation

1% 5% 10%

O 5.64 6.22 6.95

402 Stable isotope geochemistry

Answer

The d

O values have already been given.

It is left to the reader to plot the diagram.

6.0

0.703

0.705 0.707 0.709 0.711

8.0

10.0

12.0

x = 5.0

x = 1.0

O (‰)

Sr/

Crustal

contamination

Source

contamination

1:10

2:1

1:1

1:2

x = 0.2

Figure 7.24 Theoretical O–Sr isotope mixing plots. The

values show the proportion of country rock

relative to the magma. The Sr

magma

/Sr

country rock

parameter varies from 5 to 0.1. M, magma; C, crust.

After James (1981).

0.704

Marianas

arc

California

batholith

MORB

Japanese arc

0.702 0.706

O whole rock

Sr/

Sr whole rock

Figure 7.25 Example of an O–Sr isotope correlation diagram showing the Japan arcs and batholith

granites in California. After Ito and Stern (1985).

Assimilation

1% 5% 10%

Sr/

Sr 0.703 08 0.7034 0.703 94

403 Oxygen isotopes in igneous processes

7.6.2 Water–rock interaction

As we have said, isotope memory is retained if no exchange occurs after crystallization.

When th is is not the c ase, secondary isotopi c disturbances can be turnedto account. Hu gh

Tay l o r and his studentsobserved when examining variousgranitemas sifsorhydrothermal

mineral deposits that the

O isotope compositions had been disrupted after their

initial crystallization by water^rock exchanges. The calibration made on water^mineral

fractionationwas thereforeturn ed directlyto account.

Whereas the d

O values of minerals and rocks of deep origin are generally positive

(between þ5andþ8), these rocks had negative d

Ovaluesof6to7. In the same c ases,

relative fractionation as can be observed between minerals, such as quartz^potassium

feldspar fractionation, was reversed.Taylor remembered Craig’s results on thermal waters

and postulated that, rather than observing the wate rs, he was obse rving rock with which

the waters had swapped isotopes. From that point, he was able to show that the emplace-

mentofgranite plutons, especially those with associated mineral deposits, involves intense

£uid circulation inthe surrounding rock.Ofcours e, the existence ofsuch £uids wasalready

k nownbecause theygive rise toveins ofaplite and quartz pegmatitean d theyengender cer-

tain forms of mineralization around granites, but their full importance was not

u nderstood.

In a closedsystem, we canwritethe massbalance equation:

0;W

þ Rg

0;R

¼ Wg

þ Rg

where W is the mass of water and R the mass of rock, g

is the proportion of oxygen in

the water and g

the proportion of oxygen in the rock, d

0,W

and d

0,R

are the initial

compositions of water and rock, and d

and d

are the ¢nal compositions thereof.



 d

0;R

0;W

 d



;

sinc e d

and d

are related by fractionation reactions d

¼d

. This gives:



 d

0;R

0;W

ðd

 DÞ



 0:5:

Indeed,g

¼0.45 and g

¼0.89.We estimate d

0,R

from thenature ofthe rockand the catalog

of sound rock (close to þ5), and we estimate  by calibrating and estimating temperature

by fractionation among minerals.This temperature canbe compared with thetemperature

obtained by theheatbudget.

Acalculation maybe made, for example, forafeldspar with d

O ¼þ8andd

¼16

at various temperatures (Figure 7.26a). It shows that in a closed system, theW/R ratios may

be extremely variable.

404 Stable isotope geochemistry

Exercise

What is the W/R ratio of a hydrothermal system supposedly working in a closed system at

400 8C?

The initial d

O value of feldspar is þ8, that of the water determined by the meteoric

straight line is d

¼20. The 

feldspar–water

fractionation factor is 3.13 10

2

3.7. The

O of feldspar is measured as d

¼2.

Answer

The fractionation factor D ¼ 3:13  10

=ð673Þ

 3:7 ¼ 3:21 ¼ d

 d

W=R

Þ¼0:34:

Meteoric line

Measured

point

Initial composition

of the fluid

Computation of

W/R

-2

-6

-10

0.50 2 3 4 6 10 15 20 25

150 °C

TO –3.3

TO –8.9

TO –12.0

250 °C

350

°C

(W/R) closed system

rock

R = 8.5

W = –16

Figure 7.26 Variation in d

O composition and (dD, d

O) correlation diagram. (a) The variation in the d

composition of a feldspar with an initial composition d ¼þ18 is calculated as a function of (W/R) for

various temperatures, with the initial composition of water being d ¼16. (b) It is assumed the altered

rocks are represented by the blue area in the (dD, d

O) diagram. We can try to determine the initial

composition of the ﬂuid by assuming, as a ﬁrst approximation, that the dD values of the rock and water

are almost identical. The intersection between the horizontal and the (dD, d

O) correlation diagram of

rainwater gives the value of water involved in alteration. Reconstructed from several of Taylor’s papers.

405 Oxygen isotopes in igneous processes

Exercise

Let us now suppose the

ratio ¼5, that is, there is much more water. All else being equal,

what will be the d

O value of the feldspar measured?

Answer

O ¼14.54.

Theprocess describedinthe previous exercise involves adouble exchange anditis either the

water or the rock that in£uences the isotopic composition of the other depending on the

W/R ratio.

Allowing for the pointthat  varies with temperature, a whole range of scenarios canbe

generated.

Exercise

Let us pick up from where we left off in the previous exercise. Imagine an exchange between

sea water and oceanic crust whose d

¼þ5.5. The exchange occurs at

¼0.2. What will

the isotope composition of the water and rock be?

Answer

At high temperature  ¼0; maintaining

¼0.5 gives d

rock

¼þ3.64 and d

water

¼þ3.64.

Exercise

Let us imagine now that the water is driven out of the deep rock and rises to the surface and

cools to, say, 200 8C. It attains equilibrium with the country rock and its minerals. If the rock

contains feldspar, what will the isotope composition of the feldspar be?

Answer

At 100 8C, 

feldspar–water

¼10. Therefore the feldspar of the rock will have a d value of

13.92 þ14.

It can be seen from the water cycle in the previous exercise thathydrothermal circulation

reduc es the d value of deep rocks and increases the d value of surface rocks. (This is what is

observedinophiolite mas sifs.)

7.7 Paleothermometry and the water cycle:

paleoclimatology

We have just seen how hydrothermalism can be studied by combin ing information on the

isotope cycle of water and that of isotope fractionation.We are going to see how these two

e¡ects combine to give fundamental information on the evolution ofour planet and its cli-

ma te. After the initial impetus from Cesare Emiliani at Miami Universit y and Sam

Epstein attheCali forniaInstituteofTechnology,Europeanteamshavebeenthe moreactive

406 Stable isotope geochemistry

ones inthis ¢eld: for sediment paleothermometry, the teams from Cambridge and Gif-sur-

Yvette; forglacial records, thosefrom Copenhagen, Berne, Grenoble,and Saclay.

7.7.1 The two paleoclimatic records: sediments and polar ice

Carbonate paleoclimatology

In order to use oxygen isotopes as a thermometer, we must, strictly, know the d

Ovaluesof

two compounds in equi librium: water and carbonate.The formula established by Urey and

his team f or the carbonate thermometer dra w s on d

CaCO

and d

. In a ¢rst approach, the

Chicagoteam hadconsidered that dH

O,thatisthe ofseawater, wasconstantovergeologi-

caltim e andtherefore thatthe d

CaCO

measurementgave paleotemperatures directly.The dis-

covery of extreme d

O values for Antarctic ice challenged this postulate. If the amount of

Antarctic ice lost every year into the ocean varies, the d

O value of the ocean must vary

too, since this ice mayhave d

Ovaluesaslowas50. In this case, thehypothesisofconstant

O is untenable and it seems that temperatures cannotbe calculated simply. On the other

hand, ifthe dissolution ofAntarctic ice in the oceanvaries involume, thisphenomenon must

berelatedto climateandtherefore,tosomeextent,mustre£ecttheaverageglobaltemperature.

The ¢rst idea developed by Emiliani in1955 was therefore to measure the d

Ovaluesof

carbonate foraminifera in Q uaternary sediment cores for which (gl acial and interglacial)

climatic £uctuationshavelongbeenknown.Variationsin isotope compositionareobse rved

(Figure 7. 27 ) and seem to be modulated by g lacial and interglacial cycles and more

Depth in the core (cm)

Zones

Stages

500

Cold Warm Temperature (°C)

Ericson curve

for G. menardi

abundance

600

300

200

100

400

20 25 30

Isotopic curve

from

Emiliani

Figure 7.27 The ﬁrst isotope determination using 

O of Quaternary paleotemperatures by Emiliani

(1955) compared with glacial–interglacial divisions by Ericson.(

Globorot

men

rdi

is a foraminifer.)

407 Paleoclimatology

speci¢callytofollow the theoreticalpredictions oftheYugoslavastronomerMilankovitch.

Are these variations a direct e¡ect of temperature on (carbonate^water) fractionation or

arethey the e¡ectof

O dilutionbypolar i ce?Thequestion remainedunanswered.

Theformula:



¼ 16:9  4:2 d

CaCO

 d



þ 0:13 d

CaCO

 d



shows us that that two e¡ects work in the same direction. When T increas es,

dCaCO

dH

OðÞfraction ation decreases and so dCaCO

decreases if dH

O remains

constant. If, with constant local  fractionation, d

decreases for wantof polar ice then

CaCO

also declines. Ni ck Shackl eton of the University of Cambridge suggested that the

O variations measured in foram s were the result of £uctuations in the volu me of

p olar ice, a climate-relatedphenomenon.

Jean-Claude Duplessy and his colleagues in the Centre National de la Recherche

Scienti¢que (CNRS) at Gif-sur-Yvette had the idea of comparing d

O £uctuations of

surface-living (pelagic) foraminifera and bottom-dwelling (benthic) foramini fera. It is

k nownthatthetemperatureofthe deep oceanvaries l ittle around þ4 8C.

In conducting their study theyrealizedthat d

O£uctuationsofpelagic andbenthic spe-

cies were very simil ar (Figure 7.28). At most, extremely close scrutiny reveals an additional

£uctuation of 2ø inthe d

O ofpelagic species, whereas no great di¡erence appears for the

same comparison with d

C. This means, then, that th e d

O variations in foraminifera

re£ ect just as much variation in the d value of sea water as variations in local temperature.

Thesignal recorded istherefore meaningful for the global climate(Emiliani,1972).

In fact,morerecentstudieshave con¢rmedthat, for pelagic spec ies,some 50%ofthe sig-

nal re£ects a Urey-type local temperature e¡ect (remembe r that when temperature rises

O falls), above all in the temperate zones, and 50% the e¡ect of melting of the polar ice

-1

50 100

10050 200

150 200 250 300

Globigerinoides sacculifer

Planulina wuellestorfi

Depth in the core (cm)

Time (Ma)

Figure 7.28 Variation in d

O in samples of two species of foraminifer. Top: a pelagic (ocean surface)

species. Bottom: a benthic (ocean ﬂoor) species. Modiﬁed after Duplessy

.(1970).

408 Stable isotope geochemistry